PV03a-21 EMG Wearable Controller

Subscribe
Notify of
6 Comments
Inline Feedbacks
View all comments
SHEN, Shanpu(@eeshanpu)
January 19, 2022 3:34 pm

Why using PCB with SMT component can help reduce the effect from the environment? What is the advantage compared with other methods?Please clarify it?
Another comment for your poster: you font is too small to read.

WONG, Lai Shun Nelson(@lsnwong)
January 19, 2022 4:32 pm
Reply to  SHEN, Shanpu

The main purpose of using PCB with SMT component owing to the wearable design of a mobile device. At the beginning of the project, the circuit was built and test on breadboard and PCB with THT component. EMG signal is very sensitive as raw EMG signal is very weak due to high impedance on human skin. The raw signal need to be greatly amplified (100x~1000x). It is tested that circuit on breadboard will induce much noise and greatly affect the amplified signal. Circuit on PCB with THT component can significantly reduce the noise but the circuit size is not feasible for wearable application. Therefore PCB with SMT component is used additionally with 2nd order low pass filter (15kHz) to filter out noise. Sorry about the font size.

ZHANG, Jun(@eejzhang)
January 19, 2022 3:25 pm

The conclusion part mentioned “The system is capable of learning”. Can you elaborate on this aspect?

WONG, Lai Shun Nelson(@lsnwong)
January 19, 2022 3:45 pm
Reply to  ZHANG, Jun

The system uses Tensorflow training model to recognize user’s hand gesture (i.e. relax, fist, wave) from EMG signals. With more EMG signals input to the training model, it will further improve the accuracy of recognition. Later, gyroscope data will also be added to the training model to further improve the accuracy.

CAO, Xuanyu(@eexcao)
January 19, 2022 3:03 pm

What’s the main advantage of the proposed wearable controller over existing methods?

WONG, Lai Shun Nelson(@lsnwong)
January 19, 2022 4:02 pm
Reply to  CAO, Xuanyu

Most existing EMG wearable controllers are designed to wear around user’s arm, while the proposed wearable controller is designed to wear around user’s wrist. Those existing controllers are intended only to record EMG and gyroscope data while the proposed wearable controller can show the data on a LCD monitor and allows the user to interact with using integrated gesture recognition sensor (PAJ7620). It is designed to display data and interact with the user in a more comfortable and convenient way like a smart watch.